Data Visualisation and R

Matthew Johnson

Data Visualisation and R

Using R for visualising data, examples based on 2 R packages

Introduction

• trained as a freshwater ecologist
• using R and python for almost 20 years
• Data Analyst for last 10 years
• worked for Monash University and EPA

Department of Jobs Skills and Industry

• data analysis
• modelling and statistical analysis
• economic analysis
• data driven web content

Why R

Why R for Data Visualisation

R is not just for statistics

• Active Community
• Extensive capabilities

Why R for Data Visualisation

R is not just for statistics

• Reporting and web frameworks
  • rmarkdown
  • quarto
  • shiny
• charting libraries

Examples

DJSIR content is interactive, accessible and up-to-date

Tailored to client narrative

Creative Industries

Matching client branding

Creative Industries

Detailed view and download options

Victorian tourism

Headline summaries

Victorian tourism

Broad overviews

Victorian tourism

R for Data Visualisation

Charting libraries

Many choices plus R’s htmlwidgets framework

• plotly
• leaflet
• networkD3
• ggiraph

interactive charts for web and publication

configurable interactive web mapping

interactive D3 network charts

makes ggplot2 charts interactive

http://gallery.htmlwidgets.org/

Charting library examples

Creating and styling, line charts and choropleth maps

ggplot2

highcharter

ggplot2 best suited for static documents like word or PDF highcharter used for web based content

Data

Victorian unemployment data

# A tibble: 1,104 × 6
   table_no series                        series_id series_type date       value
   <chr>    <chr>                         <chr>     <chr>       <date>     <dbl>
 1 6202005  Unemployment rate ;  Persons… A8442335… Seasonally… 1978-02-01  5.96
 2 6202005  Unemployment rate ;  Persons… A8442369… Original    1978-02-01  6.58
 3 6202005  Unemployment rate ;  Persons… A8442335… Seasonally… 1978-03-01  5.72
 4 6202005  Unemployment rate ;  Persons… A8442369… Original    1978-03-01  5.94
 5 6202005  Unemployment rate ;  Persons… A8442335… Seasonally… 1978-04-01  5.49
 6 6202005  Unemployment rate ;  Persons… A8442369… Original    1978-04-01  5.50
 7 6202005  Unemployment rate ;  Persons… A8442335… Seasonally… 1978-05-01  5.64
 8 6202005  Unemployment rate ;  Persons… A8442369… Original    1978-05-01  5.71
 9 6202005  Unemployment rate ;  Persons… A8442335… Seasonally… 1978-06-01  5.53
10 6202005  Unemployment rate ;  Persons… A8442369… Original    1978-06-01  5.41
# ℹ 1,094 more rows

Data from ABS Labour Force

Victorian regional unemployment data

# A tibble: 10 × 6
   table_no series                        series_id series_type date       value
   <chr>    <chr>                         <chr>     <chr>       <date>     <dbl>
 1 6291016  >>> Melbourne - North East ;… A8459955… Original    2024-01-01   4.2
 2 6291016  >>> Melbourne - Inner ;  Une… A8459965… Original    2024-01-01   4.5
 3 6291016  >>> Ballarat ;  Unemployment… A8459966… Original    2024-01-01   7.6
 4 6291016  >>> Geelong ;  Unemployment … A8459967… Original    2024-01-01   3  
 5 6291016  >>> Hume ;  Unemployment rat… A8459967… Original    2024-01-01   2.8
 6 6291016  >>> Latrobe - Gippsland ;  U… A8459968… Original    2024-01-01   4.4
 7 6291016  >>> Melbourne - Outer East ;… A8459985… Original    2024-01-01   3.8
 8 6291016  >>> Melbourne - South East ;… A8459992… Original    2024-01-01   4.5
 9 6291016  >>> Victoria - North West ; … A8459992… Original    2024-01-01   2.6
10 6291016  >>> Melbourne - Inner East ;… A8460001… Original    2024-01-01   3.4

Data is 3 month moving average from ABS Detailed Labour force

Get spatial data

Use sf and dplyr packages to process spatial data

library(sf)

sa4 <- sf::read_sf("data/SA4_2021_AUST_GDA2020.shp") |>
  dplyr::filter(STE_NAME21 == "Victoria",
                !(SA4_CODE21 %in% c("297", "299"))) |>
  dplyr::select(SA4_NAME21)

regional_employment <- dplyr::left_join(sa4, regional_emp)

ABS digital boundaries

ggplot2

ggplot2 was first developed in 2007. Based on the “grammar of graphics” it allows you to build a chart by iteratively adding components.

ggplot2 line chart

gg_emp <- emp |>
  ggplot(aes(x = date, y = value, colour = series_type)) +
  geom_line() +
  scale_color_manual(values = c('#595959','#F8766D')) +
  labs(
    title = "Victorian unemployment rate", 
    subtitle = paste0("Monthly data since ", format(min(emp$date), "%B %Y")),
    y = "unemployment rate", x = "") 

ggplot map

gg_map <- regional_employment |>
  ggplot(aes(fill = value)) +
  geom_sf() +
  scale_fill_gradient(low = "#E2E8F4",
                      high = "#02359A") + 
  labs(title = "Victorian regional unemployment rate",
       subtitle = paste0("Regional unemployment rates for ", 
                         format(max(regional_employment$date), "%B %Y")))

ggplot map

ggplot style

gg_style <- theme(
    plot.background = element_rect(fill = "#505050", color = "#505050"),
    plot.title = element_text(color = "#fefefe"),
    plot.subtitle = element_text(color = "#fefefe"),
    plot.margin = margin(10,10,10,10),
    
    panel.background = element_rect(fill = "#505050", color = "#505050"),
    panel.grid = element_blank(),
    
    legend.background = element_rect(fill='transparent', color = "#505050"), 
    legend.box.background = element_rect(fill='transparent'), 
    legend.text = element_text(size = rel(0.8), color = "#fefefe"),
    legend.text.align = NULL,
    legend.title = element_text(hjust = 0, color = "#fefefe")
  )

https://ggplot2.tidyverse.org/reference/theme.html

ggplot styled map

gg_map + theme_void() + gg_style

highcharter

highcharter package is a wrapper for the highcharts JS library

highcharter line chart

hc_emp <- emp |>
  hchart(hcaes(x = date, y = value, group = series_type), type = "line") |>
  hc_title(text = "Victorian unemployment rate") |>
  hc_subtitle(text = paste0("Monthly data since ", 
                            format(min(emp$date), "%B %Y"))) |>
  hc_xAxis(title = list(text = NULL)) |>
  hc_yAxis(title = list(text = "unemployment rate"))

highcharter maps

aus_map <- hcmap(
  "countries/au/au-all",
  borderWidth = 0.5,
  download_map_data = TRUE) 

https://code.highcharts.com/mapdata/

highcharter map data

convert spatial data to geojson using geojsonsf

library(geojsonsf)

regional_json <- regional_employment |>
  sf::st_simplify(dTolerance = 10) |>
  geojsonsf::sf_geojson() 

highcharter map

hc_map <- highchart(type = "map") |>
  hc_add_series(
    name = "Unemployment",
    mapData = regional_json,
    data = regional_employment |> select(SA4_NAME21, value),
    value = "value",
    joinBy = "SA4_NAME21",
    borderWidth = 0.5
  ) |>
  hc_tooltip(pointFormat = "{point.SA4_NAME21}: {point.value}%") |>
  hc_colorAxis(min = min(regional_employment$value) |> floor(), 
               minColor = "#E2E8F4",
               max = max(regional_employment$value) |> ceiling(),
               maxColor = "#02359A") |>
  hc_title(text = "Victorian regional unemployment rate") |>
  hc_subtitle(text = paste0("Regional unemployment rates for ", 
                            format(max(regional_employment$date), "%B %Y")))

highcharter map theme

hc_map_theme <- highcharter::hc_theme(
      chart = list(backgroundColor = "transparent"),
      title = list(style = list(
          color = "#fefefe",
          `font-weight` =  "bold",
          `font-size` = "1.5rem",
          `background-color` = "transparent"
        )),
      subtitle = list(style = list(
          color = "#fefefe"
        )),
      colorAxis = list(labels = list(
          style = list(
            color = "#fefefe"))))

https://jkunst.com/highcharter/articles/themes.html

highcharter styled map

hc_map |> hc_size(height = 500) |> hc_add_theme(hc_map_theme)

Conclusion

• ggplot2 and highcharter are versatile and adaptive
• both have similar coding semantics
• both are easily styled with custom themes

The End

Links

This presentation

source code

slides

R resources

https://ggplot2.tidyverse.org/index.html

https://jkunst.com/highcharter/

https://r-graphics.org/

https://ggplot2-book.org/

http://gallery.htmlwidgets.org/

DJSIR Public Content

Victoria’s Economic Overview Portal

Victorian Labour Force

Victoria’s global trade

Victorian food and fibre exports

Victoria’s creative economy

Tourism in Victoria